×

You are using an outdated browser that does not fully support the intranda viewer.
As a result, some pages may not be displayed correctly.

We recommend you use one of the following browsers:

Full text

Title
Mapping without the sun
Author
Zhang, Jixian

E10: diagonal direction texture sub-image of first scale
While extracting texture structure of image, we got nine sub
images, classified three groups, first group consists of E8,
E9and E10, second group consists of E5, E6 and E7, third
group consists of E2 , E3 and E4. They express texture
structure of image on vertical direction horizontal direction
and diagonal direction on three scales. We use nine sub-image
texture structure character of image, constructing three
identified parameters separately.
A l = ASM 4/(ASM 2 + ASM 3) (2)
A 2 = ASM 7/{ASM 5 + ASM 6) (3)
A 3 = ASM 10/( ASM 8 + ASM 9) (4)
where ASM refers to equation 1
A
1 : indicator of 3 level decomposition
A 2 : indicator of 2 level decomposition
A 3 : indicator of 1 level decomposition
From equation 2, 3 and 4, the indicators of this texture index is
the ratio of average energy for texture sub-image at same scale,
so they have the invariable property to the angle of
circumrotating 90° 180° and 270° .
changjiang,
—CHANCJIANG2
changjiakl
Figure 4. code of sub-image block
The curve from the average energy of sub-image of the
changjiang river samples
Figure 5. code of sub-image block
The curve from the average energy of sub-image of house
samples
By means of the three index of texture characters it is favorable
in classification and identify.
In Quick Bird image, we cut out 64*64 image of some
kind land image for statistics character texture index 4
House
Pond
Lake
Figure 3. Texture image samples of some kinds of land
64*64 size
4. CONCLUSIONS
350
300
250
200
150
§ 100
>
<
50
0
♦ gendl
■ gend2
gend3
* gend4
Figure 6. code of sub-image block
The curve from the average energy of sub-image of infiled
samples
Figure 7. code of sub-image block
The curve from the average energy of sub-image of lake
samples
Based on 2-D lifting-based DWT 9/7wavelet filter and Gray
Level Co-occurrence Matrix method is used for texture
classification ,we can get the curve of from the average energy
of sub-image of used land samples, and texture classification
parameters of some kind of land samples.
o “
1 2
Fig
The curve from t
samples
70 r
o L - - 1 -
1 2
Fig
The curve from
samples
kind
changjiang
river
house
infiled
lake
woodlang
pond
Table 1. tex
REFERENCES
Wim Sweldens, 1
construction of bi
Anal., 3( 2), pp. If
S. BARUA, J. E. i
efficient architecti
wavelet transform:
Chen shan, qin qi-
structure in the
transform, geogrc
9.(in Chinese)
I. Daubechies.
supported wavelet
200.